We at PARIS Technologies have been audaciously proud proponents of OLAP for a long time. We named our initial product PowerOLAP. Released over 10 years, it is still considered by many as “the most Excel-user friendly” technology on the market.
You could say that we’ve stood by the online analytical processing acronym even as other vendors in the wider Business Intelligence category have tried to knock it down. There have even been ad campaigns based on the phrase—gasp—“OLAP is dead”!
And yet…the assertion continues to be made that “analytical processing” is entering a new phase. Indeed, we have our own take on that, proudly, with Olation—more on which, below.
Now there’s a new acronym in our market, HTAP. It stands for Hybrid Transactional/Analytical Processing. Is it “Online Analytical Processing” by another name? Given the close similarity (two out of four words is good—three if we consider the once-upon-a-time category HOLAP), is there a difference?
Certainly, huge advances have been made in relational database technology, and much (MUCH) more computing power is available for recording, storing and (key fact) accessing transactional data. That has allowed vendors to provide analytical results, broadly defined, to data consumers in ways that previous generations of database products just plain could not. Some popular dashboards on the market have benefited already from these advances, for example, the use of columnar database technology. They allow significantly faster retrieval of data into impressive graphical representations of key metrics. And faster processing generally allows users to interact quickly with data, at least for selecting and reordering of what in the OLAP vernacular is known as dimensions. (Interesting to note: though they may abjure the term “OLAP,” many vendors still use terms like cubes, dimensions, and the like. Hmmm.)
So Hybrid Transactional/Analytical Processing relies on newer and much more powerful, often distributed, processing: sometimes it involves a new hardware “appliance,” and it almost always requires a new software platform. Beyond this, the key point seems to be that all the technology is sited in the relational database. And so, there’s no more data replication, and new transactional information becomes part of an analytical model in as fast a time as is technologically possible.
From a semantical perspective, the previously used terms HOLAP or “ROLAP” (for relational OLAP) could work to describe these new offerings. But that would deprive vendors, and analysts, of new ways of promoting what may be significant advances in the market. That said, as ever in software, the labels shouldn’t really matter, but results should.
From our perspective, if we accept the definition of OLAP as “fast analysis of shared multidimensional information,” then these products most definitely fit the OLAP bill. We have a very broad and rigorous—at least in respect to our own products—definition of OLAP technology, too. For us, the technology should accommodate Planning applications: it should allow all-important “write back” capability to create and alter sophisticated business models related to budget planning, forecasting and essentially any kind of “what if” scenario. Also, our development efforts aim to enable users to continue working with products they use every day, so they can be far more productive and collaborative with their colleagues. (With PowerOLAP, the product we intended to work with principally was, and remains, Excel…)
With Olation, we set our sights on working with all other Business Intelligence products as well, enabling Olation to be a kind of “hub” in an overall BI strategy. Olation also is sited within the relational database itself (SQL Server, SAP HANA, etc.): the name Olation is a melding of “OLAP” and “relational.”
And so if HTAP is a new (and improved) kind of OLAP, then we are totally on board with saying that Olation already belongs to this new category—again, proudly!
Latest posts by jack@paris (see all)
- But, Does it Export to Excel? - August 31, 2015
- (Online) Analytical Processing, by Any Other Name? - December 9, 2014